The long-term goal of the proposed research is to be able to diagnose and predict periodontal disease before there are clinical symptoms of the disease. This addresses a significant health-related problem in that periodontal diseases are bacterial infections that affect over half of the US adult population and are responsible for half of all tooth loss in adults. Although it is known that there are marked differences in the microbial composition of subgingival plaque from healthy sites as compared to those sites with periodontal disease, the microbial profiles of healthy, pre-clinical sites that will progress to diseased sites are not known. The proposal is a case-controlled, longitudinal study in which progressing sites of new disease will be identified on the basis of microbial compositions or profiles. To accomplish this goal, microbial profiles will be determined by using two high-throughput technologies, namely 1) the Human Oral Microbe Identification Microarray (HOMIM) to rapidly identify the predominant oral microflora, and 2) 454 pyrosequencing to provide deep sequencing of a select subset of samples. In this grant, a biological model based on microbial compositions will be developed in order to predict periodontal disease progression. This model can be used to test new subject populations. Ultimately, the outcome of this research has great promise for bench-to-chairside applications where clinicians will be able to determine and treat those periodontal sites at risk for disease on the basis of microbial compositions.